Caroline Ellison has accepted a role at Manifund, to develop our funding platform and research how to effectively direct philanthropic dollars.
Austin Chen and Caroline Ellison, “Caroline Ellison has joined Manifund.”
This is Part 1 of a new series, Community. The effective altruist and rationalist communities are strange beasts. They are often impenetrable to outsiders and frequently misunderstood. However, understanding the structure of these communities is integral to understanding why they behave as they do. This series discusses issues surrounding the effective altruist and rationalist communities.
Today’s post explores a striking recent development. Caroline Ellison, former CEO of Alameda Research, the trading arm of FTX, has been hired by Manifund, a leading funder.
The announcement of Ellison’s hiring is admirably transparent, and Manifund executives have shown remarkable patience in responding to criticism. However, at the risk of joining the already existing pile-on, I want to discuss some reasons why this hiring should be looked on unfavorably.
Due to the recency of these developments, I prioritized speed at the possible expense of rigor in preparing this post. While I have done my best to make sure that the discussion meets a reasonable standard of rigor and accuracy, there is a higher-than-usual possibility of error, which concerns me due to the sensitivity of the material. I am open to correction and will retract and apologize for any mistaken claims.
Manifold is a social prediction platform founded by Austin Chen and Stephen and James Grugett. The platform is closely tied to the rationalist community, and to a lesser extent to the effective altruist community. Manifold has been funded by organizations including the FTX Future Fund, Survival and Flourishing Fund, and Astral Codex Ten.
Manifold hosts an annual conference, Manifest, at the Lighthaven campus run by Lightcone Infrastructure, the parent company of LessWrong. This is one of a trio of three consecutive annual events considered by many to be the central events of the rationalist social calendar. This conference has attracted significant controversy for repeatedly platforming racists.
Manifund is a funding platform run by Manifold. Manifund funds many projects broadly concerned with the rationalist and effective altruist communities. The platform runs a number of flexible funding options, and at one point even funded this blog (at a rather generous amount). I no longer receive a financial benefit from this funding, with an equivalent gift having been made to the NAACP to offset this benefit.
FTX was a cryptocurrency exchange run by Sam Bankman-Fried, reaching a peak valuation of thirty-two billion dollars. Caroline Ellison served as CEO of Alameda Research, the trading arm of FTX.
The FTX Future Fund was a charitable arm of FTX, led by figures including Will MacAskill and Nick Beckstead. The fund made over a hundred million dollars of donations, primarily to longtermist causes.
FTX collapsed after allegations of large-scale financial fraud and mismanagement. As for fraud, executives were prosecuted and convicted for misappropriating billions of dollars in customer deposits for use at Alameda Research to cover risky financial bets. As for mismanagement, FTX had a wasteful spending culture backed by an absent and eventually non-existent board of directors, and a CEO who regularly took to playing video games during business meetings.
By all appearances, there was no such entity as the FTX Future Fund. The fund was not a separately incorporated legal entity, but rather a project of FTX through which the company distributed funds, including misappropriated customer assets, to primarily longtermist projects.
Many executives received hefty sentences for their role in the financial fraud at FTX. FTX CEO Sam Bankman-Fried was sentenced to 25 years in prison on charges including wire fraud, securities fraud, and money laundering. Ryan Salame, CEO of FTX Digital Markets, was sentenced to 7.5 years.
Caroline Ellison decided not to go down with the ship. After cooperating extensively with federal investigators, including testifying directly against Bankman-Fried, Ellison was sentenced to just 2 years in prison and served only 14 months on charges including wire fraud and money laundering.
Just six months after her release from federal prison, Ellison began a pseudonymous work trial with Manifund, which was converted to a full-time job on September 11, 2026.
The announcement cited three reasons for considering Ellison: enjoyment of her writing, personal debt to the FTX Future Fund for funding Manifold, and a belief in the possibility of redemption.
The announcement justified ultimately hiring Ellison largely in terms of Manifund’s need for talent, including “useful expertise around finance and systems management.” It also cited her personal qualities.
The announcement concludes with a few words from Ellison:
I’ve said a lot about what I did wrong at Alameda/FTX—you can read what I said at my plea, during trial, and remarks at sentencing. I’m deeply sorry for what I did, and I completely understand why people may not want to work with me. But I’m very grateful to Austin for giving me a second chance, and judging me on my current work performance rather than my past. I feel very lucky not only to have a job, but to be working somewhere that I think is doing great work and has a mission I believe in.
When I started my work trial, unsure if I’d get a full-time offer at the end, I didn’t want to go through a public announcement just to potentially be let go a month later. But now that I’ve accepted a full-time offer at Manifund, I want to be upfront about it.
I’m excited to continue improving our technical infrastructure, working on operations and customer support, and posting updates on our newsletter.
What should we make of this announcement? It is, admittedly, admirably transparent. There are also some grounds for concern.
Most reputable organizations adopt an informal policy against hiring people in roles related to their previous criminal activities. Child predators are not hired in roles that require them to interact with children. Bank robbers are not hired as bank tellers. Thieves are not hired to lead corporate procurement.
This policy exists for a good reason. No matter the prospects for redemption, criminal activity of a certain type remains one of the most reliable indicators of future criminal activity of the same type. Giving convicted criminals the opportunity to commit similar crimes is widely taken to be extremely risky.
Ellison’s hiring is a striking departure from this practice. The announcement specifies that Ellison has been hired by Manifund “to develop our funding platform and research how to effectively direct philanthropic dollars.” During her work trial, she has written about topics such as regranting, impact markets, and saving for the future. Elsewhere in the post, Ellison is specifically praised for building a reconciliation tool to identify misregistered transactions on Manifold’s books.
Ellison is a convicted criminal who served time for one of the largest financial frauds in history, including nine figures worth of fraudulent charitable donations. To hire Ellison into a position at Manifund would be one thing. But to hire Ellison in a position of significant responsibility for shaping and directing charitable funds is a strong departure from the norm against hiring criminals into roles related to previous criminal activities.
After the collapse of FTX, many leading effective altruist organizations returned misappropriated customer funds that had been wrongly sent to them under the guise of charitable donations. For example, Effective Ventures returned over twenty-six million dollars in donations, amounting to 100% of funds received from the FTX Future Fund. These funds were returned because they were not legitimate charitable donations by a registered charity, but instead misappropriations of customer funds.
Rather than returning misappropriated funds, many leading rationalist organizations have refused to fully compensate the FTX estate and are engaged in ongoing litigation regarding the return of funds. For example, Lightcone Infrastructure was sued for $4.9 million of transfers from FTX-linked organizations, and appears to have settled for a significantly lower amount.
Rationalist leaders have cast this behavior as a principled stand. For example, Eliezer Yudkowsky argues on the EA Forum that:
In my possibly contrarian opinion, and speaking as somebody who I don’t think actually got any money directly from FTX Future Fund that I can recall; also speaking for myself and hastily, having not run this post past any other major leaders in EA: You are not obligated to return funding that got to you ultimately by way of FTX.
In this context, it is especially noteworthy that Manifold received over a million dollars from FTX. To my knowledge, this money has not been returned. (Chen has not yet commented publicly on the matter, citing legal concerns.)
The hiring of individuals convicted for financial crimes involving FTX should be especially concerning given Manifold’s financial relationship to FTX.
That is not how Manifold views things. In fact, the announcement of Ellison’s hiring actively praises the FTX Future Fund and lists this praise as one of the reasons why Ellison’s hiring was considered.
I felt a keen debt to the FTX Future Fund. They provided seed funding for Manifold. They funded the retreat which drew me into the EA community, and the conference where I met my wife. Much of Manifund’s work today is directly inspired by the Future Fund’s approach.
Elsewhere, Manifund CEO Austin Chen has directly defended Bankman-Fried. In this defense, Chen suggests that “the FTX blowup might be a bad judgment call, not willful fraud” and admits that Bankman-Fried “was a personal hero of mine.” Chen praises the value of loyalty and condemns the EA community’s reaction to FTX as “cowardly, coldhearted, opportunistic, bandwagon-y, two-faced and distasteful.”
In this light, Ellison’s hiring can begin to seem like a pattern of excusing misbehaviors at FTX and working directly with those involved in the FTX fraud.
One of the reasons given for hiring Ellison is the possibility of redemption. Chen writes:
I believe in redemption. Caroline has admitted her faults, worked to make creditors whole, and served her time in prison. I’d like to give her the opportunity to contribute back to the world; I hope that the world will concur.
The concern here is not with the possibility of redemption, but with how lightly redemption is earned.
Part 3 of my series on Human Biodiversity discussed Richard Hanania. We saw there that Hanania expresses a number of deeply deplorable views, including the view that people of color are genetically disposed to be less intelligent and more criminal than others, and that as a result immigration should be restricted and eugenic policies should be imposed, including forced sterilization of anyone with an IQ under 90. We saw that not only has Hanania largely failed to repudiate his racist beliefs, but also that he continues to actively peddle many of them.
This is relevant because, as we saw in Part 2 of my series on Human Biodiversity, Richard Hanania was a repeated speaker and attendee at the controversial Manifest conferences. When asked to justify Hanania’s inclusion, Chen again appealed to the possibility of redemption:
“To err is human, to forgive, divine.” I believe in second chances; he’s publicly denounced his past views and announced an intention to do better. Such admission of wrongdoing strikes me as in keeping with the spirit of making public predictions. I’ve personally (and occasionally, publicly) held radioactive opinions, and I’m grateful to my friends and communities for calling me out while continuing to work with me.
As we saw above, the possibility of redemption is being wielded incredibly lightly and forgetfully here. There is not now, and was not at the time, credible evidence of repudiation or reform, but simply a credulous, or at least willfully ignorant refusal to honestly assess the prospect of Hanania having reformed.
In a similar way, Ellison is being offered the prospect of reform at an incredibly light standard in terms of evidence and personal growth. In the very same year that Ellison was released from jail for financial crimes, Chen asks that we already excuse her due to a belief in the possibility of redemption.
Matters would be different if Ellison had spent many years publicly making up for her misdeeds and proving her capacity to work on the right side of the law. But that is not what has happened. Redemption, here, is offered quite lightly.
It may also be worth remarking on a related parallel between the Ellison and Hanania cases: ongoing financial involvement with Manifold. In Hanania’s case, we saw that Hanania’s first invitation to speak at Manifest was preceded by a forecasting tournament at Manifold which Hanania sponsored and co-organized. And in this post, we saw that Manifold received significant funding from the FTX Future Fund.
When those granted redemption are involved in significant financial relationships with the redeeming organization, this makes it harder to believe that the redemption is offered purely on the merits of the case.
One of the more striking facts about the effective altruist and rationalist communities is the free mixing of professional and personal lives. Members of these communities often live together in group houses, and there are not infrequent reports of romantic relationships bleeding into the workplace or across grantmaking relationships. Indeed, Ellison herself was romantically involved with Bankman-Fried during some of her employment at FTX.
The announcement of Ellison’s hiring at Manifund lists three reasons why Ellison was hired. The third exhibits this familiar mixing of personal and business reasoning:
Through working together, I’ve had the privilege to know Caroline as more than just an icon of FTX. I find her to be kind, humble, down to earth, hardworking, smart and witty. She cares deeply about doing right, for our users, our donors, and the world. She shares many of my sensibilities around extant problems of philanthropy; see, for example, our essay in support of for-profits. She adores Ada, my 2yo daughter (and, the feeling is clearly mutual).
While these remarks may be partly excusable as a misfired attempt to humanize Ellison, they do evidence a difficulty in separating personal relationships from business reasoning. Much of this passage, while admirable, just does not belong on a list of why someone was hired, particularly if (as it appears) they were not hired through an advertised search.
A broader concern about the Ellison hiring is simply that it is a signal of bad judgment. There was, as we saw in Section 2, a significant run of bad judgment at FTX and Alameda, including fraud, wasteful spending, and mismanagement.
There has also been evidence of poor judgment at Manifold. In particular, their annual conference, Manifest, has repeatedly platformed racists and shows little evidence of repenting and changing its way. And as we saw in Section 6, the CEO of Manifund is the author of one of the more ardent and poorly-aged defenses of Bankman-Fried’s actions at FTX.
In this vein, the Ellison hire looks like another in a string of poor judgments by Manifold. Manifold has taken a massive risk and incurred reputational loss to place a convicted financial criminal in a position of significant financial responsibility, within the very same community to which the FTX Future Fund previously channelled funds. This is not the kind of good judgment expected of an organization’s leaders.
Many effective altruists take the importance of good governance and good judgment to be among the chief lessons of the FTX scandal. Noble goals, coupled with poor judgment, can quickly go awry. If this is right, then even those broadly sympathetic to Manifund’s mission might be concerned about the impact of its leaders’ judgment on Manifund’s performance.
Effective altruists and, perhaps to a greater extent, rationalists, sometimes display a remarkable degree of disregard for the public perception of their actions.
Hiring a prominent financial criminal into a position of financial responsibility within a broader community that seeks to distance itself from the crimes at FTX does not send a good signal. It signals lack of concern for the seriousness of the crimes at FTX, and lack of desire to distance oneself from those crimes. It also signals either misunderstanding of, or lack of concern for, the nature and grounds for likely public reactions.
It is not always advisable to bow to public pressure. One of the things that makes effective altruism valuable is the willingness of the community to stake out its own perspective, even when that perspective runs contrary to received opinion. But there are limits to the degree of acceptable antagonism against public opinion, and a willingness to send costly signals against public opinion for relatively moderate organization gain is not to the credit of Manifold. They are not, here, engaged in taking a stand central to the effectiveness and survival of the movement, but taking a significant stand for modest gain. That is very hard to justify.
This post discussed the hiring of Caroline Ellison, former CEO of Alameda Research, into a position of financial responsibility at Manifund. After an introduction to Manifold, Manifund and FTX (Section 2), we looked at the announcement of Ellison’s hiring (Section 3).
Despite the admirable transparency of the announcement, we saw that this announcement raises several important concerns.
It breaks well-established conventions of not hiring convicted criminals in roles related to their crimes (Section 4). It creates significant conflict of interest given Manifold’s funding history with FTX (Section 5). It shows, for at least the second time, a cheap standard for redemption (Section 6). It involves uncomfortable mixing of personal and professional considerations (Section 7). It looks very much like one in a string of bad judgments at Manifund and FTX (Section 8). And it shows remarkable unconcern for the signals sent by this hiring (Section 9).
I don’t often write publicly about FTX. To a large extent, I am willing to respect the public narrative that leaders in the effective altruist and rationalist community were not aware of the crimes committed at FTX.
But to turn around and hire the former CEO of Alameda Research into a position of financial responsibility in the same year that she is released from prison for financial crimes is not the way for any organization or movement to credibly distance itself from the crimes at FTX. This was a poor decision that does little to quiet criticism of the community’s relationship to FTX.
Note: This essay is cross-posted from the blog Reflective Altruism written by David Thorstad. It was originally published there on September 13, 2026. The account making this post has no affiliation with Reflective Altruism or David Thorstad.
You can leave a comment on this post on Reflective Altruism here. The author will not necessarily see comments left below this cross-post.
